Re: [PATCH v5 04/16] PCI: prevent duplicate slot names

From: Alex Chiang
Date: Thu Oct 09 2008 - 01:57:10 EST


* Kenji Kaneshige <kaneshige.kenji@xxxxxxxxxxxxxx>:
>
> Thank your new patches. Very quick!!!

I'm trying to get into 2.6.28. ;)

> Though I have not reviewed/tested your patches yet (of course), I have
> one concern as I said in the e-mail soon before. Does the new one
> consider the following senario?
>
> Scenario C:
> hotplug driver(A) hotplug_driver(B)
> -------------- ----------------
> pci_create_slot(name=A, rename=1)
> pci_create_slot(name=B, rename=1)
>
> The hotplug driver (A) creates the slot with name "A". The the hotplug
> driver (B) tries to create the same slot, but wants the name "B" instead.
> In this case, hotplug driver fails to create the slot and the slot name
> should not be changed to "B" from "A".

Hm... I don't think this is a common scenario but...

int pci_hp_register(...)
{
...

pci_slot = pci_create_slot(bus, slot_nr, name, 1);
if (IS_ERR(pci_slot))
return PTR_ERR(pci_slot);

if (pci_slot->hotplug) {
dbg("%s: already claimed\n", __func__);
pci_destroy_slot(pci_slot);
return -EBUSY;
}
...
}

I could maybe move that check into pci_create_slot() instead.

struct pci_slot *pci_create_slot(...)
{
...

/*
* Get existing slot and rename if desired
*/
slot = get_slot(parent, slot_nr);
if (slot && rename) {
if ((err = slot->hotplug ? -EBUSY : 0)
|| (err = rename_slot(slot, name))) {
kobject_put(&slot->kobj);
slot = NULL;
goto err;
} else
goto out;
} else if (slot)
goto out;
...
}

Seems a little ugly to me, but maybe it's necessary?

/ac

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/